超声内镜引导细针穿刺抽吸与活检术对消化系统上皮下病变诊断价值的比较研究

Comparative diagnostic performance of endoscopic ultrasound‑guided fine‑needle aspiration versus biopsy for gastrointestinal subepithelial lesions

摘要:

目的 比较超声内镜引导细针穿刺抽吸术（fine‑needle aspiration，FNA）与细针穿刺活检术（fine‑needle biopsy，FNB）对消化系统上皮下病变诊断效能的差异。方法 回顾性分析北京友谊医院2009年1月至2024年12月期间收治的消化道上皮下病变患者。经影像学检查证实为消化道上皮下占位性病变行FNA和FNB检查。采用logistic回归分析筛选免疫组化成功的相关影响因素。使用受试者工作特征曲线以及敏感度、特异度、准确率评估FNA和FNB对消化道上皮下占位性病变的诊断效能。结果 最终纳入145例患者，FNA组56例，FNB组89例。FNB组患者免疫组化成功率［87.64%（78/89）比67.86%（38/56），χ2=7.467，P=0.006］和病理诊断成功率［93.26%（83/89）比82.14%（46/56），χ2=4.326，P=0.038］高于FNA组。对间质瘤的检出率FNB组高于FNA组［50.56%（45/89）比32.14%（18/56），χ2=5.143，P=0.023］。logistic回归分析结果显示FNB是免疫组化成功的影响因素（P=0.012，OR=2.518，95%CI：1.262~6.319）。肿瘤大小<2 cm（P=0.048，OR=0.312，95%CI：0.082~0.933）是免疫组化失败的独立风险因素。FNA和FNB对消化道上皮下占位性病变的敏感度、特异度、准确率分别为78.26%（36/46）和93.98%（78/83）、80.00%（8/10）和100.00%（6/6）、78.57%（44/56）和94.38%（84/89）。FNA和FNB对消化道上皮下占位性病变诊断的曲线下面积分别为0.791（95%CI：0.647~0.935）和0.970（95%CI：0.944~0.996），FNB对消化道上皮下占位性病变的诊断价值高于FNA（P=0.020）。结论 FNB对消化道上皮下占位性病变，尤其是间质瘤的诊断效能优于FNA。可以优先选择FNB用于消化道上皮下占位性病变的临床诊断。

Abstract:

Objective To compare the diagnostic accuracy of endoscopic ultrasound‑guided fine‑needle aspiration (FNA) and fine‑needle biopsy (FNB) for subepithelial lesions of the digestive system. Methods A retrospective study was conducted on patients with subcutaneous lesions of the digestive tract who were admitted to Beijing Friendship Hospital between January 2009 and December 2024. FNA and FNB were performed on subcutaneous space‑occupying lesions of the digestive tract that were confirmed by imaging examination. Logistic regression analysis was employed to identify factors associated with adequate immunohistochemical sampling. The diagnostic performance of FNA and FNB for gastrointestinal subepithelial lesions was evaluated using receiver operating characteristic curves, sensitivity, specificity, and accuracy. Results Among 145 patients, 56 underwent FNA and 89 underwent FNB. The FNB group showed a significantly higher immunohistochemical [87.64% (78/89) VS 67.86% (38/56), χ2=7.467, P=0.006] and pathological diagnostic success rate [93.26% (83/89) VS 82.14% (46/56), χ2=4.326, P=0.038] compared with the FNA group. The detection rate of stromal tumors in the FNB group was significantly higher than that in the FNA group [50.56% (45/89) VS 32.14% (18/56), χ2=5.143, P=0.023]. Logistic regression analysis results showed that FNB was a factor influencing the immunohistochemical success (P=0.012, OR=2.518, 95%CI: 1.262-6.319). However, tumor size <2 cm was an independent risk factor (P=0.048, OR=0.312, 95%CI: 0.082-0.933) for the failure of immunohistochemistry. Diagnostic sensitivity [93.98% (78/83) VS 78.26% (36/46)], specificity [100.00% (6/6) VS 80.00% (8/10)], and accuracy [94.38% (84/89) VS 78.57% (44/56)] of FNB were superior to FNA. The area under the curve for FNA and FNB in diagnosing subcutaneous space‑occupying lesions in the digestive tract were 0.791 (95%CI: 0.647-0.935) and 0.970 (95%CI: 0.944-0.996), respectively (P=0.020). Conclusion FNB exhibits distinct advantage over FNA in identifying subcutaneous space‑occupying lesions, particularly stromal tumor, within the digestive system. Therefore, FNB may be the preferred method for diagnosing such lesions.